US Banks Restrict AI Agents From Accessing Customer Accounts

Bank of America and Citigroup have restricted third-party artificial intelligence agents from accessing customer accounts, joining JPMorgan Chase and Charles Schwab as lenders seek to protect customer funds and prevent automated transfers to higher-yielding alternatives.

According to the UK’s Financial Times, Citi does not permit third-party AI agents or assistants to log in to customer accounts or transact directly through its digital banking platforms. Bank of America has also limited access while it considers who would be liable for payments authorized by AI tools.

Many technology companies are now developing AI assistants capable of carrying out financial tasks on users’ behalf, potentially making it easier for customers to shift money between providers.

There are, however, concerns that such AI assistants could move idle cash into other accounts offering better returns, threatening banks’ low-cost deposit bases.

JPMorgan chief executive Jamie Dimon said during the bank’s third-quarter earnings call last week that it would block outside agents attempting to access accounts using customers’ banking credentials, while supporting secure, customer-authorized data sharing through open banking providers.

“You’ve got to make sure customers aren’t hurt by it,” he said. “You don’t want rogue agents, and there could be thousands of them.”

Charles Schwab has also confirmed it is blocking unauthorized agents that attempt to log in on behalf of clients. Chief executive Rick Wurster said the firm was developing controlled ways for customers to use AI tools, including connections that allow access to information without permitting agents to move money.